As is known in the art, optical interferometric reflectometry is a powerful tool for performing non-invasive, high-resolution (˜10 μm), cross-sectional imaging of a biological or other sample, to visualize micro-structural optical properties such as reflection, absorption, scattering, attenuation, birefringence, and spectroscopic analysis. There are a number of interferometric imaging techniques that are known in the art. These techniques in general can be divided into two major categories: (i) time-domain technique, and (ii) frequency-domain technique.
Low coherence interferometry (“LCI”) is one of the time-domain techniques. This technique uses a scanning system to vary the reference arm length and acquire the interference signal at a detector. Then, the fringe pattern is demodulated to obtain the coherence envelope of the source cross correlation function. Optical coherence tomography (“OCT”) is a technique for obtaining two- or three-dimensional images using LCI. OCT is described in U.S. Pat. No. 5,321,501 issued to Swanson et al. Multiple variants of the OCT techniques have been described, but many suffer from less than optimal signal to noise ratio (“SNR”), resulting in non-optimal resolution, low imaging frame rates, and poor depth of penetration. Power usage is a factor in such imaging techniques. For example in ophthalmic uses, only a certain number of milliwatts of power are tolerable before thermal damage can occur. Thus, boosting power is not feasible to increase SNR in such environments. Nevertheless, it would be desirable to have an imaging method with superior SNR without appreciably increasing power requirements.
Insufficient SNR can also prevent the OCT technique from being used at a high frame rate which is important to avoid motion artifacts and overcome the short measurement time window available, for example, for in-vivo vascular imaging. Therefore, a way to improve SNR and imaging speed (e.g., the frame rate) is desired.
Spectral interferometry, or spectral radar, is one of the frequency-domain imaging techniques. In spectral radar, the real part of the cross spectral density of sample and reference arm light is measured with a spectrometer. Depth profile information can be encoded on the cross-spectral density modulation.
The use of spectral radar concepts to increase SNR of LCI and OCT has been described previously. This technique uses a charge coupled device (“CCD”) with a large number of pixels (an order of 1,000) to reach scan ranges on the order of a millimeter. The fast readout of the CCD device makes high-speed imaging possible.
There are, however, a number of disadvantages associated with using a CCD device. First, CCD devices are relatively expensive compared to a single-element photo-receiver. Secondly, the previously described method uses a single CCD to acquire the data. Since the charge storage capacity is limited, it requires a reduction of the reference arm power to approximately the same level as the sample arm power, giving rise to auto correlation noise on the sample arm light. In addition, since no carrier is generated, the 1/f noise will dominate the noise in this system. Thirdly, even with the short integration times of state of the art CCD technology, phase instabilities in the interferometer reduce fringe visibility of the cross spectral density modulation. This shortcoming makes the technique vulnerable to motion artifacts.
Coherent frequency-modulated continuous-wave reflectometry (C-FMCW) is another frequency domain technique known in the art. U.S. Pat. Nos. 5,956,355 and 6,160,826 issued to Swanson et al. describes an optical imaging method and apparatus using this technique. The previously described imaging method is based on using a continuously-tuned single-frequency laser as an optical source. The tuning wavelength range is required to be several tens of nanometers to achieve ranging resolution of less than 100 microns. The instantaneous linewidth of the laser must be less than approximately 0.1 nm to achieve a detection range on the order of 1.0 mm. The tuning rate should be greater than 10 kHz for high speed (e.g., video-rate) imaging. Although an external-cavity semiconductor laser can be configured to achieve mode-hop-free single-frequency tuning over several tens of nanometer, the tuning rate has been less than 1 Hz due to stringent requirement on mechanical stability. A way to overcome this speed difficulty is preferable. It would, therefore, be desirable to provide a system and method to overcome the source availability and scan speed shortcomings of conventional LCI and OCT.